DMP adoption has hit mainstream — but where is the DMP itself going?

15 Nov 2017 | By Tobias Arns

With DMP adoption — and acceptance as a digital marketer ’must have’ — continuing stronger than ever*, what is fueling this sustained and growing appetite across marketers and publishers towards the DMP? What are the 'movers and shakers' behind the evolving DMP value proposition itself that continue to capture the attention and imagination of digital marketers? In short, what are the driving forces shaping the DMP itself as the platform matures into 2020?

Below are three DMP growth and value drivers we at Cxense feel passionately about. Connect with us on Twitter (@cxense) or email (link to e.g. sales@cxense.com or info@cxense.com) we’d love to hear your thoughts and comments, and as always, we invite you to reach out for a more in-depth discussion for how the DMP can address your needs today and tomorrow alike.

Driver #1: Single Customer View

Though SCV is certainly not a novelty amongst marketing technologists, those who believe to have solved for true single customer view across channels and end user devices remain very few in numbers. Moreover, belief does not always equal reality.

With strides in cross-device identity and offline customer record matching (which started with offline data on-boarding) done by progressive DMPs, this category of enterprise software has a very strong hold on becoming the customer journey analytics module that connects the handful 'systems of record' data with the plethora of ’systems of engagement’ data, and fuses these for a genuine single customer view across both customer attributes from ’records’ as well as from ’engagement’ across channels and across devices. Already equipped with (i) real-time updating scaled ecosystem cookie match tables, deterministic and probabilistic cross-device graphs, and deterministic external offline identity (CRM file, EMS record, SSO token) matching capabilities on identity management side, combined with (ii) pixel, UTM parameter, and S2S log file based campaign performance event ingestion capabilities on customer engagement tracking side, these select progressive DMPs are perfectly positioned to providing marketers with a path to genuine SCV. It should also be noted that when instrumenting a DMP for SCV, the same architecture can be applied for SPV (Single Prospect View) when using a cross-device cookie-less identity DMP (of course such prospect profiles lack a persistent offline record association, hence remain volatile.)

Last, whilst perhaps not the most intuitive of all things, the upcoming changes in EU regulation (GDPR) is in fact providing a much needed catalyst for for true SCV to materialize at scale. As a set of data laws that force marketers to take stock of data and bridge system silos, the GDPR will inherently solve key data challenges and help them build the SCV picture they’ve always wanted. This SCV development also underscores the shift that has already taken place practically across all DMPs taking data handling from the granularity of segments to real-time updating individual 1:1 user profiles.

Driver #2: AI

With 2018 set to become The Year of AI, it is only natural that this also be viewed in the context of DMP evolution — and there is certainly plenty of good reason to be bullish on the prospects of applying machine learning and deep learning onto DMP data sets, in particular when equating DMP 1:1 user profiles with marketer SCV profiles.

Concretely, we see AI as playing a transformative role across both how audiences are created and managed in DMPs, as well as how marketers will co-ordinate their marketing activities across channels; always seeking to hit that optimally personalized next best action scenario driving conversions and CLV. For the former, we see a shift from Boolean rules based audience definition (discreet inclusion/exclusion) to machine predicted audience definition (probabilistic), wherein there are several tiers of AI application from simplistic Lookalike modeling to fusing entire data sets to fully automated audience discovery using deep learning algorithms. The foremost benefit of all three AI tiers of audience discovery for marketers is that they all reduce the need for marketers themselves to manually define which signals matter most for which activity amidst an influx of high velocity and high variance data.

The second realm of AI infusion happening in the DMP will transform how next best action scenarios are defined across marketing automation, as well which consumer engagement channels these personalized actions can be delivered in. First, we see the manually defined customer journey paths being gradually replaced by AI’s probabilistic decisioning. This is catalyzed by the inclusion of a rich new set of real-time customer engagement data harnessed by the DMP across programmatic advertising, social, paid search, and onsite personalization engagement, and fused with more stable and persistent data from marketing systems of record (yet also far less abundant and diverse). Equally, the DMP will enable automated triggering of more customer engagement channels, most notably paid search, social, and programmatic — all across devices and with frequency controls.

Driver #3: Private Data Marketplaces and Data Co-ops

The shift from segment level 3rd party data reliance to mastering one’s own 1:1 user level 1st party data continues relentlessly, as marketers seek to unify their data silos. This trend is further strengthened by the DMP powered quest for SCV, and the tightening EU data regulation (GDPR) in 2018.

There is however a very real limit to how far most marketers alone can get relying only on their own 1st party accessible data assets. Quality alone does not fully compensate for reach. Complementing 1st party data with readily available 3rd party segment level data does not offer much solace either, as opacity continues to plague the 3rd party data markets, and as segment membership level information falls short of most needs beyond simple campaign targeting.

In stead, marketers are pouring in over 2nd party data partnership opportunities; both to access qualified, transparent, and trusted peer segments, as well as to access each others’ 1:1 level audience data alike. Whilst the former is already materializing in robust growth for Private Data Marketplaces, also called 2nd Party Data Marketplaces or Peer Data Marketplaces, the rise of mutual AI ambitions is driving another class of privileged and private data sharing: data co-ops for 1:1 level user data. Here, we see a plethora of activity and excitement with DMP customers eager to open up their 1st party data (using the term broadly here, not restricting its meaning to on-boarded offline data) for peers, in exchange for mutual benefit — often in the form of improved algorithmic audiences or cross-device graphs.

Feel free to connect with us on via email if you'd like to connect for a more in-depth discussion on how the Cxense DMP can address your needs today!

*Forrester predicts the category to see "robust 43% annual growth over the 2015 to 2021 period” (source: Forrester Ad Technology (Data Management Platforms) Forecast, 2016 To 2021 (US). April 12, 2016)